Determinants of house prices in Seoul: A quantile regression approach

Introduction:

A house can be regarded as a bundle of utility-bearing attributes that are valued by consumers. These attributes are characterized by their physical inflexibility, durability and spatial fixity such that different combinations of them can produce a heterogeneous good. In the real estate literature the house price is defined as a function of a bundle of inherent attributes (e.g. flat size, age, and floor level), neighbourhood characteristics (e.g. scenic view), accessibility (e.g.: to metro station and school) and environmental quality (e.g. fresh air or natural beauty) that yield utility or satisfaction to homebuyers. In particular, a hedonic price model by ordinary least squares (OLS) has been utilized in several studies to model the relationship between a set of housing attributes and price.

To address this problem, a quantile regression method is adopted that models the relation between a set of explanatory variables and each quantile of house prices. By estimating the changes in a specific quantile produced by a one-unit change in an explanatory variable, the quantile regression can give a more comprehensive picture of the effect of explanatory variables on house prices and the difference in the level of the effect at the same time.

Literature review:

The quantile regression model introduced by Koenker and Bassett is more flexible than OLS. Quantile regression allows examination of more comprehensive pictures of different house price levels. The quantile regression is based on the minimization of weighted absolute deviations for estimating conditional quantile (percentile) functions. For the median, symmetrical weights are used, while asymmetrical weights are employed for all other quantiles. While the traditional OLS regression estimates conditional mean functions, quantile regression can be employed to explain the determinants of the dependent variable at any point on the distribution of the dependent variable.

Data on House Prices:

For this study, three counties in Seoul were selected: Kangnam, Songpa, and Nowon. The reason for choosing these counties for this study is that Kangnam is a top premium submarket in Seoul, Songpa is generally considered to be the second, while Nowon is a moderate and relatively cold submarket. Therefore, adopting these three counties as a sample is considered to be adequate and appropriate in representing various aspects of the Seoul housing market. In fact, Seoul, the capital city of Korea, is divided into two areas by the Han River, which runs from east to west through the middle of the city. These two areas include Kangnam (south of the river) and Kangbuk (north of the river).

Both are located in the Seoul metropolitan area: Kangnam is a relatively new region consisting of 11 counties characterized by better living conditions including superior housing interiors, amenities, and, in particular, a favorable educational environment, while Kangbuk is older, with moderately changing house prices behaviour.

Empirical results:

The empirical analysis is conducted by estimating Equation with seven quantiles, the 5th, 10th, 25th, 50th, 75th, 90th, and 95th. This allows examination of the impact of the explanatory variables at different points of the housing price distribution, presents the empirical results obtained by the hedonic pricing model using the traditional OLS method. Panel A in presents the estimation results of the quantile regressions, goodness of fit measure and their diagnostic statistics. The slope equality test and symmetric quantile test results in show that coefficients differ across quantile values and that the conditional quantiles are not identical.

Conclusions:

This paper empirically estimates how specific quantiles of house prices in Seoul respond differently to a one-unit change in the hedonic characteristics of the house. An analysis of each county is conducted and the effects of the hedonic attributes on house prices compared between three regions of Seoul. A total data sample of 3459 court auction cases traded from 2006 to 2012 in three different counties of Seoul is used, Kangnam, Songpa and Nowon. To check the robustness of the estimation results and to analyze the impact of the 2008 financial crisis on the housing market, the study period is divided into two sub-periods, being before and after the crisis.

Also Read: The Iranian Housing System under the Islamic Republic Government (1990–2021)